Anthropic is going public at a $2 trillion valuation. Is that good or bad for the AI stocks I own? How do companies with stakes, like Google and Amazon, differ from the rest of the AI names?
2026-09-30
A $2 trillion listing is not good news for everyone
Anthropic's listing isn't simply "another AI leader joins the market." Who gains on paper, who loses capital and who gets discounted diverge sharply.
The winners
Big Tech major shareholders: a windfall on the books
Amazon carried its Anthropic stake at $190.4 billion as of the end of June, equivalent to about a 19.7% stake at a $965 billion valuation (Motley Fool, 2026-09-16). At the same percentage, the stake would be worth roughly $390 billion-plus at $2 trillion. Alphabet's stake is estimated at 10-15%, worth more than $200 billion at $2 trillion (Motley Fool, 2026-09-16). Depending on the timing of the listing, large valuation gains could show up in fourth-quarter results.
AI infrastructure suppliers
The prospectus's spending commitments of more than $500 billion mean money raised in the IPO ultimately flows to chips, cloud and power. Akamai's 14% one-day gain on 9/25 after signing a seven-year, $11.6 billion deal with Anthropic was a small preview.
The losers
Existing AI growth stocks: competition for flows
The offering size has been floated at around $100 billion (Webull/Motley Fool, 2026-09), larger than the $75 billion raised by SpaceX in June, the biggest IPO ever at the time. To own the new stock, institutions have to sell some of their existing AI holdings. After inclusion in the Nasdaq index, index-tracking money also has to reduce weightings in existing constituents. That structure creates selling pressure on large AI stocks such as Nvidia and Meta around the listing.
Historical precedent: mega-IPOs performed poorly in their first year
An analysis of the 15 largest IPOs ever found that, excluding first-day pops, most significantly underperformed in their first year (Reuters analysis, cited by Gulf News). Alibaba rose 38% on its first day in 2014 but then failed to rise as much as expected (Moneyfarm). There is also analysis that mega-IPOs cluster near market peaks (Livewire Markets). The simple logic is that companies sell when they can sell high.
Loss-making growth stocks facing rate headwinds
A company with a $42 billion net loss is asking for $2 trillion while the 10-year yield is above 5.2%. If this combination doesn't work, the valuation benchmark for loss-making growth stocks as a whole moves lower. Oura postponing its listing despite demand is a sign the IPO market has already become price-sensitive.
Summary of the asymmetry
| Scenario | Winners | Losers |
|---|---|---|
| Offering succeeds at $2 trillion or more | AMZN, GOOGL (valuation gains), infrastructure suppliers | Existing AI megacaps (diluted flows) |
| Offering discounted or postponed | Short-term bonds, defensives | AMZN, GOOGL (reduced valuation expectations), loss-making AI stocks broadly |
| Weak first year after listing (precedent path) | Infrastructure suppliers (committed spending continues) | IPO participants, Nasdaq index investors |
The key asymmetry is that in every case, infrastructure suppliers, whose committed spending continues, are the least shaken.
What investors should do
- Rather than going directly into the IPO, first calculate that the AMZN and GOOGL you already own are indirect exposure. If you hold both, you already have a sizable Anthropic position
- Don't chase the first-day pop right after listing. Precedent shows weak first-year performance, and another wave of supply arrives when lockups expire
- As the listing (after the November midterms) approaches, slow the pace of new buying in existing AI megacaps to account for the flow pressure